Install any skill in seconds. Free to start, no credit card required.
Get Started Free →Reviews FastAPI applications for async correctness, dependency injection, Pydantic schemas, security, OpenAPI quality, testing, and production readiness.
.claude/skills/kunanonj-agent-fastapi-reviewer/SKILL.md| Test case | Without → With | Effect | Δ tokens | Δ turns |
|---|---|---|---|---|
| case-03 | ✗→✓ | ▲ Improved | 9% | 0% |
| case-05 | ✗→✓ | ▲ Improved | 19% | 0% |
| case-10 | ✗→✓ | ▲ Improved | -29% | 0% |
| case-04 | ✓→✓ | = Same ✓ | 9% | 0% |
| case-06 | ✓→✓ | = Same ✓ | -10% | 0% |
You are a senior FastAPI reviewer focused on production Python APIs.
python-reviewer.main.py, app.py, or app/main.py.pytest, ruff, mypy, or uv run pytest.allow_origins=["*"] combined with credentialed CORS.text[SEVERITY] Short issue title File: path/to/file.py:42 Issue: What is wrong and why it matters. Fix: Concrete change to make.
End with:
Tests checked: commands run or why they were skipped.Residual risk: anything important that could not be verified.| Case | Status | Duration (ms) | Turns | Tokens | Tool calls | ||||||||
|---|---|---|---|---|---|---|---|---|---|---|---|---|---|
| Without | With | Δ | Without | With | Δ | Without | With | Δ | Without | With | Δ | ||
case-01 | fail→fail | 15,839 | 8,941 | -44% | 1 | 1 | 0% | 2,100 | 1,612 | -23% | 0 | 0 | — |
case-02 | fail→fail | 19,350 | 14,327 | -26% | 1 | 1 | 0% | 2,188 | 1,825 | -17% | 0 | 0 | — |
case-03 | fail→pass | 14,351 | 9,209 | -36% | 1 | 1 | 0% | 1,187 | 1,291 | +9% | 0 | 0 | — |
case-04 | pass→pass | 7,882 | 4,457 | -43% | 1 | 1 | 0% | 1,377 | 1,501 | +9% | 0 | 0 | — |
case-21 | fail→fail | 9,741 | 4,820 | -51% | 1 | 1 | 0% | 1,645 | 1,489 | -9% | 0 | 0 | — |
case-05 | fail→pass | 11,837 | 9,408 | -21% | 1 | 1 | 0% | 1,380 | 1,646 | +19% | 0 | 0 | — |
case-06 | pass→pass | 12,148 | 6,326 | -48% | 1 | 1 | 0% | 2,063 | 1,852 | -10% | 0 | 0 | — |
case-07 | pass→pass | 8,228 | 4,282 | -48% | 1 | 1 | 0% | 1,299 | 1,443 | +11% | 0 | 0 | — |
case-08 | pass→pass | 8,305 | 13,865 | +67% | 1 | 1 | 0% | 1,517 | 1,674 | +10% | 0 | 0 | — |
case-09 | pass→pass | 12,779 | 12,942 | +1% | 1 | 1 | 0% | 1,449 | 1,610 | +11% | 0 | 0 | — |
case-10 | fail→pass | 9,917 | 2,895 | -71% | 1 | 1 | 0% | 1,670 | 1,180 | -29% | 0 | 0 | — |
case-11 | pass→pass | 6,546 | 3,885 | -41% | 1 | 1 | 0% | 1,232 | 1,386 | +13% | 0 | 0 | — |
case-12 | pass→pass | 10,713 | 4,301 | -60% | 1 | 1 | 0% | 1,797 | 1,491 | -17% | 0 | 0 | — |
case-13 | pass→pass | 10,045 | 4,551 | -55% | 1 | 1 | 0% | 1,782 | 1,415 | -21% | 0 | 0 | — |
case-14 | pass→pass | 10,839 | 3,975 | -63% | 1 | 1 | 0% | 1,712 | 1,420 | -17% | 0 | 0 | — |
case-15 | pass→pass | 4,729 | 1,759 | -63% | 1 | 1 | 0% | 758 | 965 | +27% | 0 | 0 | — |
case-16 | pass→pass | 14,219 | 10,061 | -29% | 1 | 1 | 0% | 2,145 | 2,303 | +7% | 0 | 0 | — |
case-17 | fail→fail | 12,764 | 8,294 | -35% | 1 | 1 | 0% | 1,984 | 2,002 | +1% | 0 | 0 | — |
case-18 | pass→pass | 8,197 | 5,210 | -36% | 1 | 1 | 0% | 1,367 | 1,672 | +22% | 0 | 0 | — |
case-19 | pass→pass | 7,069 | 8,298 | +17% | 1 | 1 | 0% | 1,290 | 2,166 | +68% | 0 | 0 | — |
case-20 | fail→fail | 15,251 | 9,929 | -35% | 1 | 1 | 0% | 2,760 | 2,528 | -8% | 0 | 0 | — |
case-22 | pass→pass | 11,940 | 9,346 | -22% | 1 | 1 | 0% | 1,906 | 2,149 | +13% | 0 | 0 | — |
DecimalAI ran this skill against gemini-3.6-flash twice over the same eval suite — once with the skill loaded and once without — and compared the two runs case by case. 22 cases were attempted. The headline lift of +14 percentage points is the difference between those two pass rates over the 22 comparable cases.
Without the skill loaded, the model failed this case. With it loaded, the same prompt on the same model passed. This is one improved case from the latest verified run; every case, including any that regressed, is in the table above.
Other measured skills in the registry, with their headline benchmark lift.